In recent years, development of autonomous vehicles has been, and increasingly is, growing rapidly. The concept of autonomous driving relates to that the vehicle, at least to some extent, is driven without human interaction. That is, the vehicle may have an automation level of e.g. 0%<automation level≦100%, wherein 0% may represent that the vehicle is driven by a driver only, and 100% that the vehicle is driven completely autonomously. When having an automation level anywhere between 0% and 100%, the vehicle may autonomously perform some actions, as e.g. keeping a suitable distance to the vehicle ahead, while the driver may perform other actions, as e.g. overtaking another vehicle when appropriate. The closer to 100%, the more actions are performed autonomously by the vehicle.
Autonomous vehicles, also commonly known as autonomously driven vehicles, driverless vehicles, self-driving vehicles, or robot vehicles, are known to sense their surrounding with such techniques as e.g. radar, lidar, GPS and/or computer vision. Advanced control systems may interpret sensory information to identify appropriate paths, as well as obstacles. For instance, an autonomous vehicle may update its map based on sensory input, allowing the vehicle to keep track of its position even when conditions change or when it enters uncharted environments. Notwithstanding the foregoing, it is challenging to provide accurate vehicle positioning.
WO 2013/149149, for instance, relates to identifying a driven lane on a map and to improve a vehicle position estimate. There is suggested estimating a location of the vehicle using global navigation satellite system, and determining a closest mapped road to the estimated location of the vehicle. Furthermore, there is suggested that light detection and ranging (lidar) and camera systems are used in detection of, or ranging to, distinct features in the environment and that for vehicle positioning, the sensors may be used to detect street signs, buildings, or lane markings in the immediate vicinity of the vehicle.
However, although WO 2013/149149 discloses a solution enabling for refining the estimated location of the vehicle, the challenge of providing a vehicle position estimate with high accuracy, remains.